When a computer system suffers from disasters such as a fire, flood, an earthquake, or a war, if no effective measure is taken, cloud applications may be interrupted and data may be lost in the entire computer system, which brings about a great loss. A redundancy technology can effectively deal with impact on the computer system from an ambient environment, and particularly, a disastrous event. In the case of a disaster, the redundancy technology can back up data at a production site to a redundancy site, which ensures security of user data and high availability of the system.
A virtualization technology integrates a plurality of operating systems into one high-performance server, which maximizes the utilization of all resources of a hardware module, and can implement more cloud applications with less investment. This can not only simplify a system architecture, but also reduce resource management difficulty, so that the technology is widely applied in computer systems. At present, the prior art mainly uses the following manner to implement redundancy for cloud applications running on a virtualized computer system.
An administrator determines a virtual machine included in a cloud application needing redundancy, that is, the administrator specifies a virtual machine needing redundancy, then queries a virtualization module at a production site to determine a storage unit used by the virtual machine needing redundancy, and configures a corresponding storage unit at a redundancy site for the virtual machine needing redundancy, so that the redundancy site can replicate data in the storage unit used by the virtual machine needing redundancy to the corresponding storage unit at the redundancy site.
When redundancy needs to be performed for the production site, a virtual machine is created at the redundancy site, and corresponding configuration is performed for the created virtual machine by means of manual operations, for example, a corresponding network is configured, so that storage and a network of the virtual machine needing redundancy can be recovered at the redundancy site, so as to complete virtualized redundancy of the production site.
The prior art recovers a virtual machine needing redundancy at a production site, and an administrator specifies a virtual machine for which backup and recovery are to be implemented for redundancy. Because a cloud application is deployed at multiple virtual machines, if a part of virtual machines are not specified when the administrator specifies the virtual machine needing redundancy, the part of virtual machines will not be recovered, and finally, the cloud application cannot be completely recovered. In addition, during a recovery process of the virtual machines in the prior art, networks of the virtual machines need to be configured manually one by one, causing the recovery process of multiple virtual machines to be slow and at low efficiency.